Systems And Methods For Processing Images Related To Boundaries

Abstract

Systems and methods are provided for processing images related to boundaries. An example computer-implemented method includes accessing an image data set, which includes multiple images of an agricultural field; masking the image data set based on one or more criteria; calculating, by a computing device, one or more composites from the image data set; generating, by the computing device, using a model, a segmentation of each of the one or more composites, based on the one or more composites; and combining the segmentation(s) for the one or more composites into a field boundary for the agricultural field.

What is claimed is: 
     
         1 . A computer-implemented method for processing images related to boundaries, the computer-implemented method comprising:
 accessing an image data set, which includes multiple images of an agricultural field;   masking the image data set based on one or more criteria;   calculating, by a computing device, one or more composites from the image data set;   generating, by the computing device, using a model, a segmentation of each of the one or more composites, based on the one or more composites; and   combining the segmentation(s) for the one or more composites into a field boundary for the agricultural field.   
     
     
         2 . The computer-implemented method of  claim 1 , wherein the multiple images include multiple images over a time of interest (TOI), which includes multiple seasons of a calendar year. 
     
     
         3 . The computer-implemented method of  claim 1 , wherein the one or more criteria includes a pixel validation and a blur effective metric, relative to at least one threshold. 
     
     
         4 . The computer-implemented method of  claim 1 , wherein the one or more composites include an RGB mean composite, a RGB median composite, and/or RGB standard deviation composite; and/or
 wherein the one or more composites include an NVDI in the 5 th  percentile composite, an NDVI in the 95% percentile composite, and/or a delta composite between the 5 th  percentile and in the 95 th  percentile.   
     
     
         5 . The computer-implemented method of  claim 1 , wherein the model includes a segment anything model. 
     
     
         6 . The computer-implemented method of  claim 5 , further comprising defining one or more prompts for the image data set; and
 wherein generating the segments for each of the one or more composites is further based on the defined one or more prompts.   
     
     
         7 . The computer-implemented method of  claim 6 , further comprising accessing crop-specific land cover data for the image data set; and
 wherein the one or more prompts includes a bounding-box prompt.   
     
     
         8 . The computer-implemented method of  claim 1 , wherein combining the segmentation(s) includes combining the segmentation(s) based on one or more rules related to under-segmentation, over-segmentation, and/or consistency of boundary demarcations across the one or more composites. 
     
     
         9 . A system for processing images related to boundaries, the system comprising at least one computing device configured to:
 access an image data set, which includes multiple images of an agricultural field;   mask the image data set based on one or more criteria;   calculate one or more composites from the image data set;   generate, using a model, a segmentation of each of the one or more composites, based on the one or more composites; and   combine the segmentation(s) for the one or more composites into a field boundary for the agricultural field.   
     
     
         10 . The system of  claim 9 , wherein the multiple images include multiple images over a time of interest (TOI), which includes multiple seasons of a calendar year. 
     
     
         11 . The system of  claim 9 , wherein the one or more criteria includes a pixel validation and a blur effective metric, relative to at least one threshold. 
     
     
         12 . The system of  claim 9 , wherein the one or more composites include an RGB mean composite, a RGB median composite, and/or RGB standard deviation composite; and/or wherein the one or more composites include an NVDI in the 5 th  percentile composite, an NDVI in the 95% percentile composite, and/or a delta composite between the 5 th  percentile and in the 95 th  percentile. 
     
     
         13 . The system of  claim 9 , wherein the model includes a segment anything model. 
     
     
         14 . The system of  claim 13 , wherein the at least one computing device is further configured to define one or more prompts for the image data set; and
 wherein the at least one computing device is configured, in order to generate the segments for each of the one or more composites, to generate the segments further based on the defined one or more prompts.   
     
     
         15 . The system of  claim 14 , wherein the at least one computing device is further configured to access crop-specific land cover data for the image data set; and
 wherein the one or more prompts includes a bounding-box prompt.   
     
     
         16 . The system of  claim 9 , wherein the at least one computing device is configured, in order to combine the segmentation(s), to combine the segmentation(s) based on one or more rules related to under-segmentation, over-segmentation, and/or consistency of boundary demarcations across the one or more composites. 
     
     
         17 . One or more non-transitory computer-readable media including executable instructions for processing images related to boundaries, which when executed by at least one processor, cause the at least one processer to:
 access an image data set, which includes multiple images of an agricultural field;   mask the image data set based on one or more criteria;   calculate one or more composites from the image data set;   generate, using a model, a segmentation of each of the one or more composites, based on the one or more composites; and   combine the segmentation(s) for the one or more composites into a field boundary for the agricultural field.   
     
     
         18 . The one or more non-transitory computer-readable media of  claim 17 , wherein the multiple images include multiple images over a time of interest (TOI), which includes multiple seasons of a calendar year; and
 wherein the one or more criteria includes a pixel validation and a blur effective metric, relative to at least one threshold.   
     
     
         19 . The one or more non-transitory computer-readable media of  claim 17 , wherein the one or more composites include an RGB mean composite, a RGB median composite, and/or RGB standard deviation composite; and/or
 wherein the one or more composites include an NVDI in the 5 th  percentile composite, an NDVI in the 95% percentile composite, and/or a delta composite between the 5 th  percentile and in the 95 th  percentile.   
     
     
         20 . The one or more non-transitory computer-readable media of  claim 17 , wherein the model includes a segment anything model; and
 wherein the executable instructions, when executed by the at least one processor, further cause the at least one processer to:   access crop-specific land cover data for the image data set;   define one or more bounding-box prompts for the image data set; and   generate the segments further based on the defined one or more bounding-box prompts.